Introduction to R

R is a popular open-source programming language and environment for statistical computing and graphics. In this section, we will cover basic R syntax, functions, data structures, and data manipulation using the tidyverse.

R Syntax and Basic Functions

# Arithmetic operations
2 + 3

## [1] 5

4 * 5

## [1] 20

6 / 2

## [1] 3

# Variables
x <- 10
y <- 20
x + y

## [1] 30

# Functions
mean(c(1, 2, 3, 4, 5))

## [1] 3

Data Structures in R

R has several data structures for organizing and storing data. Understanding these data structures is essential for effective data manipulation and analysis.

Vectors: A one-dimensional array of elements, all of the same data type. Vectors are commonly used for performing arithmetic operations and storing sequences of values.

v <- c(1, 2, 3, 4, 5)

Matrices: A two-dimensional array of elements, organized in rows and columns, and all of the same data type. Matrices are useful for linear algebra operations and storing tabular data with a fixed number of rows and columns.

m <- matrix(1:9, nrow = 3, ncol = 3)
print(m)

##      [,1] [,2] [,3]
## [1,]    1    4    7
## [2,]    2    5    8
## [3,]    3    6    9

Data frames: A two-dimensional table where each column can have a different data type. Data frames are the most common data structure for storing and manipulating tabular data in R, especially for statistical analysis.

df <- data.frame(
  name = c("Alice", "Bob", "Charlie"),
  age = c(25, 30, 35),
  city = c("New York", "San Francisco", "Los Angeles")
)
print(df)

##      name age          city
## 1   Alice  25      New York
## 2     Bob  30 San Francisco
## 3 Charlie  35   Los Angeles

Lists: A versatile data structure that can store elements of different data types and structures, including vectors, matrices, and other lists. Lists are useful for organizing and storing hierarchical or nested data.

l <- list(
  name = "Alice",
  age = 25,
  city = "New York"
)
print(l)

## $name
## [1] "Alice"
## 
## $age
## [1] 25
## 
## $city
## [1] "New York"

Each data structure has its strengths and applications, and choosing the right one depends on the specific needs of your data manipulation and analysis tasks.

Loading the Tidyverse and ggplot2

# Install tidyverse and ggplot2 if not already installed
if (!requireNamespace("tidyverse", quietly = TRUE)) {
  install.packages("tidyverse")
}
if (!requireNamespace("ggplot2", quietly = TRUE)) {
  install.packages("ggplot2")
}

# Load the tidyverse and ggplot2
library(tidyverse)

## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
## ✔ ggplot2 3.3.6      ✔ purrr   0.3.5 
## ✔ tibble  3.1.8      ✔ dplyr   1.0.10
## ✔ tidyr   1.2.1      ✔ stringr 1.4.1 
## ✔ readr   2.1.3      ✔ forcats 0.5.2 
## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag()    masks stats::lag()

library(ggplot2)

Reading Data from a File

R supports a wide variety of file formats for reading and writing data. Some common formats include:

CSV (Comma-Separated Values)
TSV (Tab-Separated Values)
Excel (XLS, XLSX)
JSON (JavaScript Object Notation)
XML (eXtensible Markup Language)
SQLite (a lightweight database format)

The tidyverse packages, such as readr and readxl, provide functions for reading data from many of these formats. In this workshop, we’ll focus on reading data from CSV, TSV, or Excel files.

Reading CSV Data

# Read data from a CSV file
data_csv <- read_csv("your_data_file.csv")

## New names:
## Rows: 29 Columns: 22
## ── Column specification
## ──────────────────────────────────────────────────────── Delimiter: "," chr
## (11): imdb, title, test, clean_test, binary, code, director, director_ge... dbl
## (11): ...1, year, budget, domgross, intgross, budget_2013$, domgross_201...
## ℹ Use `spec()` to retrieve the full column specification for this data. ℹ
## Specify the column types or set `show_col_types = FALSE` to quiet this message.
## • `` -> `...1`

# Display the first few rows of the data
head(data_csv)

## # A tibble: 6 × 22
##    ...1  year imdb      title  test  clean…¹ binary budget domgr…² intgr…³ code 
##   <dbl> <dbl> <chr>     <chr>  <chr> <chr>   <chr>   <dbl>   <dbl>   <dbl> <chr>
## 1     0  2013 tt1711425 21 &a… nota… notalk  FAIL   1.3 e7  2.57e7  4.22e7 2013…
## 2     1  2012 tt1343727 Dredd… ok-d… ok      PASS   4.5 e7  1.34e7  4.09e7 2012…
## 3     2  2013 tt2024544 12 Ye… nota… notalk  FAIL   2   e7  5.31e7  1.59e8 2013…
## 4     3  2013 tt1272878 2 Guns nota… notalk  FAIL   6.1 e7  7.56e7  1.32e8 2013…
## 5     4  2013 tt0453562 42     men   men     FAIL   4   e7  9.50e7  9.50e7 2013…
## 6     5  2013 tt1335975 47 Ro… men   men     FAIL   2.25e8  3.84e7  1.46e8 2013…
## # … with 11 more variables: `budget_2013$` <dbl>, `domgross_2013$` <dbl>,
## #   `intgross_2013$` <dbl>, `period code` <dbl>, `decade code` <dbl>,
## #   director <chr>, director_gender <chr>, genre <chr>, rating <dbl>,
## #   country <chr>, language <chr>, and abbreviated variable names ¹clean_test,
## #   ²domgross, ³intgross

Reading TSV Data

Remove # from second and last row to run example

# Read data from a TSV file
# data_tsv <- read_tsv("your_data_file.tsv")

# Display the first few rows of the data
# head(data_tsv)

Reading Excel Data

# Install readxl package if not already installed
if (!requireNamespace("readxl", quietly = TRUE)) {
  install.packages("readxl")
}

# Load the readxl package
library(readxl)

Remove # from second and last row to run example

# Read data from an Excel file
# data_excel <- read_excel("your_data_file.xlsx", sheet = "Sheet1")

# Display the first few rows of the data
# head(data_excel)

Replace "your_data_file.csv", "your_data_file.tsv", and "your_data_file.xlsx" with the appropriate file paths for your dataset. For the Excel file, also specify the correct sheet name using the sheet parameter in the read_excel() function.

High-Level Overview of the Data with Tidyverse Pipes

Tidyverse pipes, represented by the %>% <ctrl+shift+M> symbol, allow you to chain together multiple functions in a clear and readable manner. Pipes take the output of one function and use it as the input for the next function, making it easy to follow the sequence of data transformations.

In the following example, we’ll use pipes to perform a series of operations on our dataset: 1. Group the data by a specific category (group_by) 2. Calculate summary statistics for each group (summarize) 3. Sort the resulting summary by a specific statistic (arrange)

# Summarize the data
data_summary <- data %>%
  group_by(YourCategory) %>%
  summarize(
    mean_value = mean(YourVariable, na.rm = TRUE),
    n = n()
  ) %>%
  arrange(desc(mean_value))

head(data_summary)

## # A tibble: 5 × 3
##   YourCategory mean_value     n
##   <fct>             <dbl> <int>
## 1 B                  50.7    25
## 2 A                  50.0    20
## 3 C                  49.6    11
## 4 E                  49.1    24
## 5 D                  46.5    20

In the above code, replace YourCategory and YourVariable with the appropriate column names from your dataset. This will give a high-level overview of the data, summarizing it by the specified category and calculating the mean of the specified variable. The use of pipes makes it easy to understand the sequence of transformations applied to the data.

Task:

Use the cell below to repeat the steps we have done so far. Download data from here (if you haven’t done so already). Read the data to R and try tidyverse summarize, count, etc. Find more options from data-wrangling-cheatsheet

btd <- read_tsv('bechdel_test - bechdel_test.tsv')

## New names:
## Rows: 1794 Columns: 22
## ── Column specification
## ──────────────────────────────────────────────────────── Delimiter: "\t" chr
## (11): imdb, title, test, clean_test, binary, code, director, director_ge... dbl
## (11): ...1, year, budget, domgross, intgross, budget_2013$, domgross_201...
## ℹ Use `spec()` to retrieve the full column specification for this data. ℹ
## Specify the column types or set `show_col_types = FALSE` to quiet this message.
## • `` -> `...1`

Plotting with ggplot2

The ggplot2 package in R follows a modular paradigm based on the “Grammar of Graphics.” This modular approach allows users to build complex plots by combining simple components or layers. Each layer represents a specific element of the plot, such as data, aesthetics, geoms, scales, and themes.

Data: This is the foundation of any plot. You specify the dataset you want to visualize.
Aesthetics: Aesthetics are the visual properties of the plot, such as x and y position, color, size, shape, and transparency. You map variables in the dataset to these aesthetics, creating a relationship between the data and the plot elements.
Geoms: Geoms (short for geometric objects) are the actual plot elements, such as points, lines, and bars. Different geoms represent different types of plots, like scatter plots, line plots, or bar plots. You can add multiple geoms to a single plot to create complex visualizations.
Scales: Scales control how data values are mapped to aesthetic properties. They define the transformation and mapping of data values to visual properties, such as colors, sizes, or shapes. You can adjust scales to customize the appearance of the plot.
Themes: Themes control the non-data aspects of the plot, such as the background, gridlines, axis labels, and legend. You can customize the plot’s appearance by changing its theme.

In the ggplot2 modular paradigm, you start by specifying the data and the aesthetics, then add geoms, scales, and themes as needed. This layer-by-layer approach allows you to create a wide range of plots by combining and customizing these components.

Here’s an example that demonstrates the modular paradigm:

library(ggplot2)

# Define data and aesthetics
plot <- ggplot(data = mtcars, aes(x = wt, y = mpg, color = hp))
plot

# Add geom
plot <- plot + geom_point()
plot

# Adjust scale
plot <- plot + scale_color_continuous(low = "blue", high = "red")
plot

# Apply theme
plot <- plot + theme_minimal()

# Display the final plot
plot

In this example, we first define the data and aesthetics, then add a point geom, adjust the color scale, and finally apply a minimal theme. The result is a scatter plot that shows the relationship between car weight and miles per gallon, with points colored according to horsepower.

Aesthetics in ggplot2

Aesthetics are the visual properties of the elements in a plot. They help convey the underlying patterns and relationships in the data. In ggplot2, you map variables from your dataset to aesthetics to create a relationship between the data and the plot elements. Common aesthetics include x and y position, color, size, shape, group, and transparency.

Mapping Aesthetics

Let’s create a scatter plot with the year variable mapped to the x-axis, the revenue variable mapped to the y-axis, and the bechdel_result variable mapped to the color and shape aesthetics:

# btd <- read_tsv('bechdel_test - bechdel_test.tsv')
# Create a scatter plot with aesthetics mapped to variables
scatter_plot <- ggplot(btd, aes(x = year, y = rating, color = factor(genre))) +
  geom_point()

scatter_plot

## Warning: Removed 3 rows containing missing values (geom_point).

Modifying Aesthetic Properties and Overwriting in Geoms

You can modify aesthetic properties directly within a geom. This allows you to make specific adjustments to the appearance of individual plot elements. Let’s create a scatter plot with larger points and a custom transparency:

# Modify the size and transparency of points within the geom
scatter_plot_independent_aes <- ggplot(btd, aes(x = year, y = rating)) +
  geom_point(size = 3, alpha = 0.3, color='gold')

scatter_plot_independent_aes

## Warning: Removed 3 rows containing missing values (geom_point).

scatter_plot_data_related_aes <- ggplot(btd, aes(x = year, y = rating)) +
  geom_point(size = 3, alpha = 0.3, aes(color = factor(genre)))

scatter_plot_data_related_aes

## Warning: Removed 3 rows containing missing values (geom_point).

In this example, we have increased the size of the points and made them semi-transparent by setting the size and alpha arguments within the geom_point() function, respectively. By modifying aesthetics directly within a geom, you can overwrite the initial mappings created by the aes() function and gain more control over the appearance of your plot elements.

Exploring Different Geom Selections

In this section, we will demonstrate various geom selections using simple dummy data. Each geom represents a specific type of plot, and their applicability depends on the type of data (numeric vs. categorical) and the relationship you want to visualize.

Scatterplot (geom_point)

A scatterplot displays the relationship between two numeric variables by plotting points at their respective x and y coordinates. It’s useful for visualizing trends, patterns, or outliers in the data.

# Use the mtcars dataset as an example
head(mtcars)

##                    mpg cyl disp  hp drat    wt  qsec vs am gear carb
## Mazda RX4         21.0   6  160 110 3.90 2.620 16.46  0  1    4    4
## Mazda RX4 Wag     21.0   6  160 110 3.90 2.875 17.02  0  1    4    4
## Datsun 710        22.8   4  108  93 3.85 2.320 18.61  1  1    4    1
## Hornet 4 Drive    21.4   6  258 110 3.08 3.215 19.44  1  0    3    1
## Hornet Sportabout 18.7   8  360 175 3.15 3.440 17.02  0  0    3    2
## Valiant           18.1   6  225 105 2.76 3.460 20.22  1  0    3    1

# Create a scatterplot of mpg (miles per gallon) vs. wt (weight)
scatter_plot <- ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point()

scatter_plot

Line Plot (geom_line)

A line plot connects data points with lines to visualize the relationship between two numeric variables. It’s useful for showing trends over time or any continuous variable.

# Use the pressure dataset as an example
head(pressure)

##   temperature pressure
## 1           0   0.0002
## 2          20   0.0012
## 3          40   0.0060
## 4          60   0.0300
## 5          80   0.0900
## 6         100   0.2700

# Create a line plot of temperature vs. pressure
line_plot <- ggplot(pressure, aes(x = temperature, y = pressure)) +
  geom_line()

line_plot

Bar Plot (geom_bar) and Column Plot (geom_col)

A bar plot displays the frequency or count of categorical data, while a column plot displays the value of a numeric variable for each category. Both are useful for visualizing relationships between categorical and numeric variables.

# Use the diamonds dataset as an example
head(diamonds)

## # A tibble: 6 × 10
##   carat cut       color clarity depth table price     x     y     z
##   <dbl> <ord>     <ord> <ord>   <dbl> <dbl> <int> <dbl> <dbl> <dbl>
## 1  0.23 Ideal     E     SI2      61.5    55   326  3.95  3.98  2.43
## 2  0.21 Premium   E     SI1      59.8    61   326  3.89  3.84  2.31
## 3  0.23 Good      E     VS1      56.9    65   327  4.05  4.07  2.31
## 4  0.29 Premium   I     VS2      62.4    58   334  4.2   4.23  2.63
## 5  0.31 Good      J     SI2      63.3    58   335  4.34  4.35  2.75
## 6  0.24 Very Good J     VVS2     62.8    57   336  3.94  3.96  2.48

# Create a bar plot of diamond counts by cut
bar_plot <- ggplot(diamonds, aes(x = cut)) +
  geom_bar()

bar_plot

# Create a column plot of average price by cut
column_plot <- ggplot(diamonds, aes(x = cut, y = price)) +
  geom_col()

column_plot

Histogram (geom_histogram)

A histogram groups numeric data into bins and displays the frequency of observations in each bin. It’s useful for visualizing the distribution of a numeric variable.

# Create a histogram of the mpg variable from the mtcars dataset
histogram_plot <- ggplot(mtcars, aes(x = mpg)) +
  geom_histogram(binwidth = 2)

histogram_plot

Box Plot (geom_boxplot)

A box plot displays the distribution of a numeric variable across different categories. It’s useful for comparing distributions and identifying outliers within categorical groups.

# Create a box plot of price by cut for the diamonds dataset
box_plot <- ggplot(diamonds, aes(x = cut, y = price)) +
  geom_boxplot()

box_plot

These are just a few examples of the many geoms available in ggplot2. By selecting the appropriate geom for your data, you can create informative and visually appealing plots that effectively communicate the relationships within your dataset.

Position Adjustments and Adding Multiple Layers

Jittering Points (position = “jitter”)

In scatter plots, data points can sometimes overlap, making it difficult to discern individual observations. To address this issue, you can apply a position adjustment, such as jitter, which slightly moves the points in a random direction to reduce overlap.

# Create a scatter plot of the diamonds dataset with jittered points
jittered_plot <- ggplot(diamonds, aes(x = cut, y = price)) +
  geom_point(position = "jitter", alpha = 0.5) +
  theme_minimal()

jittered_plot

In this example, the points are jittered to reduce overlap, making it easier to see the distribution of observations within each cut category.

Adding Multiple Layers (Geoms)

You can add multiple geoms to a single plot to create more complex visualizations. For example, you can combine a scatter plot with a smoothed line to show the overall trend, and add text labels to annotate specific data points.

# Use the mtcars dataset as an example
head(mtcars)

##                    mpg cyl disp  hp drat    wt  qsec vs am gear carb
## Mazda RX4         21.0   6  160 110 3.90 2.620 16.46  0  1    4    4
## Mazda RX4 Wag     21.0   6  160 110 3.90 2.875 17.02  0  1    4    4
## Datsun 710        22.8   4  108  93 3.85 2.320 18.61  1  1    4    1
## Hornet 4 Drive    21.4   6  258 110 3.08 3.215 19.44  1  0    3    1
## Hornet Sportabout 18.7   8  360 175 3.15 3.440 17.02  0  0    3    2
## Valiant           18.1   6  225 105 2.76 3.460 20.22  1  0    3    1

# Create a scatter plot of mpg (miles per gallon) vs. wt (weight) with a smoothed line
scatter_plot_smooth <- ggplot(mtcars, aes(x = wt, y = mpg, label = rownames(mtcars))) +
  geom_point() +
  geom_smooth(method = "loess", se = FALSE, linetype = "dashed", color = "blue") +
  theme_minimal()

scatter_plot_smooth

## `geom_smooth()` using formula 'y ~ x'

# Annotate specific data points with text labels
scatter_plot_annotate <- scatter_plot_smooth +
  geom_text(data = subset(mtcars, mpg > 30 | wt > 5), 
            size = 3, hjust = -0.2, vjust = 0.5, 
            mapping = aes(label=rownames(subset(mtcars, mpg > 30 | wt > 5)))) +
  annotate(geom = 'text', x = 5, y = 30, label="Text that I can add here", color="blue")

scatter_plot_annotate

## `geom_smooth()` using formula 'y ~ x'

In this example, we first created a scatter plot of mpg vs. wt and added a smoothed line using geom_smooth. Then, we annotated specific data points with text labels using geom_text. By combining multiple geoms, you can create more informative and visually appealing plots.

Task:

Use the cell below to repeat the steps we have done so far. Generate a plot where you show diamonds data by cut and price. Use geom_violin, what does it show and how to interpret it? Add geom_point to there as well, make it very light (transparent), use position jitter. Find more options from data-wrangling-cheatsheet

Scaling Numeric and Discrete Variables

Scaling variables allows you to transform their range, making it easier to visualize data that spans multiple orders of magnitude or to compare multiple variables with different units.

Scaling Numeric Variables

For numeric variables, you can use scale_x_continuous() and scale_y_continuous() to adjust the scales of the x and y axes. Let’s create a scatter plot of intcross vs. rating and scale the axes using log transformations:

# Create a scatter plot of intcross vs. rating with log-scaled axes
scatter_plot_scaled <- ggplot(btd, aes(x = rating, y = intgross)) +
  geom_point() +
  scale_y_continuous(trans = "log10")

scatter_plot_scaled

## Warning: Removed 14 rows containing missing values (geom_point).

In this example, we used log transformations to scale the intcross variable. This can help reveal patterns in the data that might not be apparent when using the original scales.

Scaling Discrete Variables

For discrete variables, you can use scale_x_discrete() and scale_y_discrete() to modify the order or appearance of the categories. Let’s create a box plot of genre vs. rating and reorder the genres by median rating:

# Calculate the median rating for each genre
genre_median <- btd %>%
  group_by(genre) %>%
  summarize(median_rating = median(rating, na.rm = TRUE)) %>%
  arrange(median_rating) %>% 
  filter(!is.na(genre)) %>%
  filter(!genre%in%c('Family', 'Musical', 'Romance', 'Thriller'))

# Create a box plot of genre vs. rating with reordered categories
box_plot_scaled <- ggplot(btd, aes(x = genre, y = rating)) +
  geom_boxplot() +
  scale_x_discrete(limits = genre_median$genre) +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

box_plot_scaled

## Warning: Removed 8 rows containing missing values (stat_boxplot).

In this example, we reordered the genre categories based on the median rating. This can help highlight differences in the distribution of rating across genres and make it easier to compare them.

Scaling variables can enhance the readability of your plots and reveal hidden patterns in your data. By applying appropriate transformations to numeric and discrete variables, you can create more effective visualizations.

Color Scaling

Color scaling can be used to visualize an additional variable in a plot, adding an extra dimension of information. In this example, we will create a scatter plot of budget vs. intgross, with the color of the points representing the rating.

# Create a scatter plot of budget vs. intgross with color-scaled ratings
scatter_plot_color <- ggplot(btd, aes(x = budget, y = intgross, color = rating)) +
  geom_point(alpha = 0.4) +
  scale_x_continuous(labels = scales::comma, trans = "log10") +
  scale_y_continuous(labels = scales::comma, trans = "log10") +
  scale_color_continuous(low = "blue", high = "red") +
  theme_minimal()

scatter_plot_color

## Warning: Removed 11 rows containing missing values (geom_point).

In this example, we used scale_color_continuous() to adjust the color scale based on the rating variable. The points are colored from blue (low rating) to red (high rating), providing an extra layer of information on top of the relationship between budget and intgross.

By applying color scaling, you can create richer visualizations that display more information and enhance the understanding of your data.

Refactoring and Reordering Factorial Data

Factor variables are categorical variables that can take a limited number of distinct values. In ggplot2, you can refactor and reorder factor variables to enhance your visualizations.

Refactoring

Refactoring involves changing the levels of a factor variable. You can use the forcats package (part of the tidyverse) to refactor variables. The fct_recode() function can be used to recode the levels of a factor.

# Load the required packages
# library(forcats)

# Create a new column with a refactored clean_test variable
btd_refactored <- btd %>%
  mutate(clean_test_refactored = fct_recode(clean_test,
                                            "pass" = "ok",
                                            "fail" = "nowomen",
                                            "fail" = "notalk",
                                            "fail" = "men",
                                            "N/A" = "dubious"))

head(btd_refactored)

## # A tibble: 6 × 23
##    ...1  year imdb      title  test  clean…¹ binary budget domgr…² intgr…³ code 
##   <dbl> <dbl> <chr>     <chr>  <chr> <chr>   <chr>   <dbl>   <dbl>   <dbl> <chr>
## 1     0  2013 tt1711425 21 &a… nota… notalk  FAIL   1.3 e7  2.57e7  4.22e7 2013…
## 2     1  2012 tt1343727 Dredd… ok-d… ok      PASS   4.5 e7  1.34e7  4.09e7 2012…
## 3     2  2013 tt2024544 12 Ye… nota… notalk  FAIL   2   e7  5.31e7  1.59e8 2013…
## 4     3  2013 tt1272878 2 Guns nota… notalk  FAIL   6.1 e7  7.56e7  1.32e8 2013…
## 5     4  2013 tt0453562 42     men   men     FAIL   4   e7  9.50e7  9.50e7 2013…
## 6     5  2013 tt1335975 47 Ro… men   men     FAIL   2.25e8  3.84e7  1.46e8 2013…
## # … with 12 more variables: `budget_2013$` <dbl>, `domgross_2013$` <dbl>,
## #   `intgross_2013$` <dbl>, `period code` <dbl>, `decade code` <dbl>,
## #   director <chr>, director_gender <chr>, genre <chr>, rating <dbl>,
## #   country <chr>, language <chr>, clean_test_refactored <fct>, and abbreviated
## #   variable names ¹clean_test, ²domgross, ³intgross

In this example, we refactored the clean_test variable by changing the levels from “yes” and “no” to “pass” and “fail”.

Reordering

Reordering involves changing the order of the levels of a factor variable. You can use the fct_reorder() function to reorder the levels of a factor based on the values of another variable.

# Create a new column with reordered genre based on the median intgross
btd_reordered <- btd_refactored %>%
  mutate(genre_reordered = fct_reorder(genre, intgross, .fun = median, .desc = TRUE)) %>% 
  filter(!(is.na(genre) | genre%in%c('Family', 'Musical', 'Romance', 'Thriller')))

head(btd_reordered)

## # A tibble: 6 × 24
##    ...1  year imdb      title  test  clean…¹ binary budget domgr…² intgr…³ code 
##   <dbl> <dbl> <chr>     <chr>  <chr> <chr>   <chr>   <dbl>   <dbl>   <dbl> <chr>
## 1     0  2013 tt1711425 21 &a… nota… notalk  FAIL   1.3 e7  2.57e7  4.22e7 2013…
## 2     1  2012 tt1343727 Dredd… ok-d… ok      PASS   4.5 e7  1.34e7  4.09e7 2012…
## 3     2  2013 tt2024544 12 Ye… nota… notalk  FAIL   2   e7  5.31e7  1.59e8 2013…
## 4     3  2013 tt1272878 2 Guns nota… notalk  FAIL   6.1 e7  7.56e7  1.32e8 2013…
## 5     4  2013 tt0453562 42     men   men     FAIL   4   e7  9.50e7  9.50e7 2013…
## 6     5  2013 tt1335975 47 Ro… men   men     FAIL   2.25e8  3.84e7  1.46e8 2013…
## # … with 13 more variables: `budget_2013$` <dbl>, `domgross_2013$` <dbl>,
## #   `intgross_2013$` <dbl>, `period code` <dbl>, `decade code` <dbl>,
## #   director <chr>, director_gender <chr>, genre <chr>, rating <dbl>,
## #   country <chr>, language <chr>, clean_test_refactored <fct>,
## #   genre_reordered <fct>, and abbreviated variable names ¹clean_test,
## #   ²domgross, ³intgross

In this example, we reordered the levels of the genre variable based on the median value of intgross for each genre. This can be useful for creating plots where the categories are ordered meaningfully.

Now let’s create a bar plot of the reordered genres with the refactored clean_test variable:

# Create a bar plot of the refactored clean_test and reordered genre variables
bar_plot_reordered <- ggplot(btd_reordered, aes(x = genre_reordered, fill = clean_test_refactored)) +
  geom_bar(position = "dodge") +
  labs(x = "Genre (reordered)", fill = "Clean Test (refactored)") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

bar_plot_reordered

In this plot, we used the refactored clean_test variable and the reordered genre variable to create a more informative visualization. By refactoring and reordering factor variables, you can enhance the readability and effectiveness of your plots.

Enhancing Plots with Titles, Axis Labels, Limits, and Themes

Adding titles, axis labels, limits, and customizing the theme can make your plots more informative and visually appealing. In this section, we will demonstrate how to enhance a plot using these features.

Let’s create a scatter plot of budget vs. intgross and use the rating column for color scaling:

scatter_plot_example <- ggplot(btd, aes(x = budget, y = intgross, color = rating)) +
  geom_point(alpha = 0.7) +
  scale_x_continuous(labels = scales::comma, trans = "log10") +
  scale_y_continuous(labels = scales::comma, trans = "log10") +
  scale_color_continuous(low = "blue", high = "red") +
  theme_minimal()

scatter_plot_example

## Warning: Removed 11 rows containing missing values (geom_point).

Adding Title and Axis Labels

You can add a main title, subtitle, and axis labels using the labs() function:

scatter_plot_labeled <- scatter_plot_example +
  labs(
    title = "Budget vs. International Gross",
    subtitle = "Colored by Rating",
    x = "Budget (log scale)",
    y = "International Gross (log scale)",
    color = "Rating"
  )

scatter_plot_labeled

## Warning: Removed 11 rows containing missing values (geom_point).

Setting Axis Limits

To set the limits for the x and y axes, you can use the xlim() and ylim() functions:

scatter_plot_limited <- scatter_plot_labeled +
  xlim(100000000, 3e+8) +
  ylim(100000000, 3e+8) + 
  geom_point(mapping = aes(size = rating), alpha=0.4)

## Scale for 'x' is already present. Adding another scale for 'x', which will
## replace the existing scale.

## Scale for 'y' is already present. Adding another scale for 'y', which will
## replace the existing scale.

scatter_plot_limited

## Warning: Removed 1735 rows containing missing values (geom_point).
## Removed 1735 rows containing missing values (geom_point).

Customizing the Theme

You can customize the appearance of your plot by modifying the theme. The theme() function allows you to change various aspects of the plot, such as text size, font, background colors, and grid lines:

scatter_plot_custom_theme <- scatter_plot_limited +
  theme(
    plot.title = element_text(size = 18, face = "bold"),
    plot.subtitle = element_text(size = 14),
    axis.title = element_text(size = 12, face = "bold"),
    axis.text = element_text(size = 10),
    panel.background = element_rect(fill = "white"),
    panel.grid.major = element_line(color = "gray", linetype = "dashed", size = 0.5),
    panel.grid.minor = element_line(color = "gray", linetype = "dotted", size = 0.25)
  )

scatter_plot_custom_theme

## Warning: Removed 1735 rows containing missing values (geom_point).
## Removed 1735 rows containing missing values (geom_point).

By adding titles, axis labels, limits, and customizing the theme, you can create more informative and visually appealing plots that effectively communicate your data’s story. See cheatsheet for additional themes.

Exporting and Saving Plots with ggsave

ggsave() is a convenient function for saving your ggplot2 plots in various file formats, such as PNG, PDF, SVG, or TIFF. Using ggsave(), you can easily save high-quality versions of your plots for use in reports, presentations, or publications.

Let’s create a scatter plot of budget vs. intgross with the rating column for color scaling:

scatter_plot_example <- ggplot(btd, aes(x = budget, y = intgross, color = rating)) +
  geom_point(alpha = 0.7) +
  scale_x_continuous(labels = scales::comma, trans = "log10") +
  scale_y_continuous(labels = scales::comma, trans = "log10") +
  scale_color_continuous(low = "blue", high = "red") +
  labs(
    title = "Budget vs. International Gross",
    subtitle = "Colored by Rating",
    x = "Budget (log scale)",
    y = "International Gross (log scale)",
    color = "Rating"
  ) +
  theme_minimal()

scatter_plot_example

## Warning: Removed 11 rows containing missing values (geom_point).

To save this plot as a high-quality PNG image, you can use the ggsave() function:

# Save the plot as a PNG file
ggsave(
  filename = "scatter_plot_example.png",
  plot = scatter_plot_example,
  width = 8,
  height = 5,
  dpi = 300
)

## Warning: Removed 11 rows containing missing values (geom_point).

In this example, we saved the scatter_plot_example plot as a PNG file with a width of 8 inches, a height of 5 inches, and a resolution of 300 dots per inch (DPI). You can adjust the width, height, and dpi parameters to control the size and quality of the saved image. Here plot = argument may be optional – if plot is not assigned it will by default take the last active plot.

To save the plot in a different file format, you can change the file extension in the filename parameter. For example, to save the plot as a PDF, you can use:

# Save the plot as a PDF file
ggsave(
  filename = "scatter_plot_example.pdf",
  plot = scatter_plot_example,
  width = 8,
  height = 5
)

## Warning: Removed 11 rows containing missing values (geom_point).

By using ggsave(), you can easily export and save your ggplot2 plots in a variety of file formats to share or include in your documents.

Combining Multiple Plots with patchwork

The patchwork package allows you to easily combine multiple ggplot2 plots into a single layout. This can be useful for comparing different visualizations side by side or creating more complex visualizations.

First, let’s create some example plots using the mtcars dataset:

# Load the required packages
library(ggplot2)
library(patchwork)

# Create a scatter plot of mpg vs. wt
scatter_plot <- ggplot(mtcars, aes(x = wt, y = mpg)) +
  geom_point(aes(color = cyl)) +
  labs(title = "Miles per Gallon vs. Weight", x = "Weight (1000 lbs)", y = "Miles per Gallon", color = "Cylinders") +
  theme_minimal() + 
  theme(title = element_text(size = 12, face = "bold"), plot.subtitle = element_text(size = 10))

# Create a bar plot of the number of cars per number of cylinders
bar_plot <- ggplot(mtcars, aes(x = factor(cyl))) +
  geom_bar(aes(fill = factor(cyl))) +
  labs(title = "Number of Cars per Number of Cylinders", x = "Number of Cylinders", y = "Count", fill = "Cylinders") +
  theme_minimal() + 
  theme(plot.title = element_text(size = 8, face = "bold"))

# Create a box plot of mpg per number of cylinders
box_plot <- ggplot(mtcars, aes(x = factor(cyl), y = mpg)) +
  geom_boxplot(aes(fill = factor(cyl))) +
  labs(title = "Miles per Gallon per Number of Cylinders", x = "Number of Cylinders", y = "Miles per Gallon", fill = "Cylinders") +
  theme_minimal() + 
  theme(plot.title = element_text(size = 8, face = "bold"))

Now, let’s use the patchwork package to combine these plots:

# Combine the plots using patchwork
combined_plot <- scatter_plot + bar_plot + box_plot + plot_layout(ncol = 1)
combined_plot

In this example, we combined the three plots into a single column layout. You can adjust the layout by changing the ncol and nrow parameters in the plot_layout() function.

To collect the legends and add a global title, subtitle, and caption, you can use the plot_annotation() function:

# Collect legends and add global title, subtitle, and caption
combined_plot_annotated <- scatter_plot / (bar_plot + box_plot) +
  plot_annotation(
    title = "Exploring the mtcars Dataset",
    subtitle = "Scatter plot, bar plot, and box plot",
    caption = "Data source: mtcars",
    theme = theme(plot.title = element_text(size = 13, face = "bold"), plot.subtitle = element_text(size = 11)),
    tag_levels = 'A'
  ) +
  plot_layout(guides = "collect")

combined_plot_annotated

By using the patchwork package, you can combine multiple ggplot2 plots into a single layout, making it easier to compare and present your visualizations.

Task:

Now it’s your time to shine. Let you imagination fly and explore the bechdel test dataset. Come up with your own visualisation. Use as many plots, as many colours and themes as you see fit. Start building it gradually, this makes finding and identifying errors easier. You can even take your own data and explore it. But try to get to a publishing ready image (sans figure legend).

Find more options from data-wrangling-cheatsheet.

Introduction to R and ggplot2

Priit Adler

2023-05-02

Introduction to R

R Syntax and Basic Functions

Data Structures in R

Loading the Tidyverse and ggplot2

Reading Data from a File

Reading CSV Data

Reading TSV Data

Reading Excel Data

High-Level Overview of the Data with Tidyverse Pipes

Task:

Plotting with ggplot2

Aesthetics in ggplot2

Mapping Aesthetics

Modifying Aesthetic Properties and Overwriting in Geoms

Exploring Different Geom Selections

Scatterplot (geom_point)

Line Plot (geom_line)

Bar Plot (geom_bar) and Column Plot (geom_col)

Histogram (geom_histogram)

Box Plot (geom_boxplot)

Position Adjustments and Adding Multiple Layers

Jittering Points (position = “jitter”)

Adding Multiple Layers (Geoms)

Task:

Scaling Numeric and Discrete Variables

Scaling Numeric Variables

Scaling Discrete Variables

Color Scaling

Refactoring and Reordering Factorial Data

Refactoring

Reordering

Enhancing Plots with Titles, Axis Labels, Limits, and Themes

Adding Title and Axis Labels

Setting Axis Limits

Customizing the Theme

Exporting and Saving Plots with ggsave

Combining Multiple Plots with patchwork

Task: